---
name: analytics
description: "Define business-relevant marketing measurement, inspect data quality, report performance, and recommend the next decision without overstating attribution."
---

# Marketing analytics

Connect marketing activity to useful customer actions and business decisions. Start with the objective and what the business can actually observe.

## Define the measurement question

Read the business goal, conversion action, sales process, and timeframe. Identify the decision the report should inform. Separate reach, engagement, inquiries, qualified opportunities, customers, and repeat business.

Define terms before calculating. A click is not an inquiry, an inquiry is not a qualified opportunity, and booked revenue is not cash or profit. Record how the business identifies a suitable lead and a completed sale.

## Inspect the available data

Use authorized reports, exports, website events, or sales records. Record source, coverage dates, time zone, currency, filters, and access limits. Check missing values, duplicates, changed definitions, and test activity.

If the business has no website, use an appropriate booking, platform, or manual inquiry record. Do not require a website analytics product merely to track a non-website journey.

## Plan collection

Use the [measurement worksheet](references/measurement-worksheet.md). Name the meaningful event, trigger, source, owner, and verification method. Add campaign tags where supported and document their meaning.

Review an existing implementation before proposing tag changes. Check the real event flow when changes are authorized. A reporting connection does not prove that the website records the right action.

Avoid placing private customer data or credentials in event parameters, tracking links, or shared reports. Apply requirements relevant to the actual business and location rather than inventing a universal compliance checklist.

## Analyze and explain

Calculate ratios with explicit numerators and denominators. Use comparable time windows and describe missing or small samples. Distinguish an observed association from evidence of a causal change.

When presenting a growth scenario, label its assumptions and show the calculation. Do not reuse another business's sales values, margins, or conversion rates as a forecast.

For return calculations, state which costs and revenue are included and how revenue was attributed. Avoid a precise return claim when attribution or costs are incomplete.

## Recommend a decision

Report what happened, the strongest explanation supported by evidence, reasonable alternatives, and what to do next. Tie recommendations to the business's workload, budget, and customer quality.

Deliver a concise scorecard with the date range, goal, key measures, data quality limitations, and a specific next action. Preserve the source records or reproducible calculation where practical. Do not manufacture success to justify the plan.